Sidescan Sonar Imagery of the Winter Marginal Ice Zone Obtained from an AUV

P. Wadhams Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Cambridge, and Scottish Association for Marine Science, Dunstaffnage Marine Laboratory, Oban, Argyll, United Kingdom

Search for other papers by P. Wadhams in
Current site
Google Scholar
PubMed
Close
,
J. P. Wilkinson Scottish Association for Marine Science, Dunstaffnage Marine Laboratory, Oban, Argyll, United Kingdom

Search for other papers by J. P. Wilkinson in
Current site
Google Scholar
PubMed
Close
, and
A. Kaletzky Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Cambridge, United Kingdom

Search for other papers by A. Kaletzky in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

The first Arctic under-ice sidescan sonar imagery from an autonomous underwater vehicle (AUV) has been obtained in the winter marginal ice zone of the East Greenland Current at 73°00′N, 11°47′W, using a Maridan Martin 150 vehicle operated from R/V Lance. First-year, multiyear, brash, and frazil ice can be discriminated, with the underside of multiyear ice appearing smooth as compared to the rough underside detected by submarine-borne sidescan in the Arctic Basin, implying downstream bottom melt. The ice draft profile was obtained from the vertical part of the sidescan beam, and the probability density function of ice thickness was derived and found to agree well with upward sonar results from this region obtained in 1987 from a British submarine.

Corresponding author address: Prof. Peter Wadhams, Scottish Association for Marine Science, Dunstaffnage Marine Laboratory, Oban, Argyll PA37 1QA, United Kingdom. Email: peter.wadhams@sams.ac.uk

Abstract

The first Arctic under-ice sidescan sonar imagery from an autonomous underwater vehicle (AUV) has been obtained in the winter marginal ice zone of the East Greenland Current at 73°00′N, 11°47′W, using a Maridan Martin 150 vehicle operated from R/V Lance. First-year, multiyear, brash, and frazil ice can be discriminated, with the underside of multiyear ice appearing smooth as compared to the rough underside detected by submarine-borne sidescan in the Arctic Basin, implying downstream bottom melt. The ice draft profile was obtained from the vertical part of the sidescan beam, and the probability density function of ice thickness was derived and found to agree well with upward sonar results from this region obtained in 1987 from a British submarine.

Corresponding author address: Prof. Peter Wadhams, Scottish Association for Marine Science, Dunstaffnage Marine Laboratory, Oban, Argyll PA37 1QA, United Kingdom. Email: peter.wadhams@sams.ac.uk

1. Introduction

Recent evidence of changes in the Arctic indicate that the sea ice cover is undergoing a significant thinning (Rothrock et al. 1999; Wadhams and Davis 2000) and retreat (e.g., Bjørgo et al. 1997). General circulation models using greenhouse gas forcing predict that the Arctic ice cover will continue to diminish because of global warming and may become seasonal by the 2080s. At the same time, it is possible that the current ice shrinkage is a response to oceanic and atmospheric changes associated with the Arctic Oscillation (AO) (Thompson and Wallace 1998), and that these will reverse when the AO itself changes phase. To resolve this question, it is essential to monitor the Arctic ice cover on a seasonal and interannual basis. This is easy from the point of view of extent, because of the availability of satellites, but difficult from the point of view of thickness.

Current methods of ice thickness monitoring have been reviewed by Wadhams (2000). Essentially, the only direct satellite-borne technique is the radar altimeter, which measures freeboard and which has not yet been fully validated in comparative experiments; other satellite-based techniques using synthetic aperture radar (SAR) or passive microwave involve inference from other measured parameters. Airborne techniques (laser altimetry for freeboard, electromagnetic sounding for thickness) are expensive for obtaining data over large areas, while through-ice techniques (hole drilling, surface sounding) are purely local. This leaves under-ice mapping as the most readily available and commonly used technique, involving the use of upward-looking sonar from moorings or from submarines. Once again, moorings offer data only at fixed locations, even though these may be critical choke points (e.g., Fram Strait), so only submarines have offered true synoptic ice-thickness mapping. The continued availability of submarines (U.S. and British) is therefore essential to the task of monitoring Arctic ice thickness through the present period of rapid change. Since the end of the Cold War, however, the deployment of British submarines in the Arctic has become more sporadic, and the U.S. civilian Scientific Ice Expeditions (SCICEX) program, which also produced many valuable data on Arctic sea ice from submarines, has been reduced in scope.

Given the probable continued shortage of submarine availability, the use of autonomous underwater vehicles (AUVs) under sea ice is clearly an option that needs to be pursued vigorously, as AUVs offer the only platform that will be definitely available for scientifically controlled missions. The vehicles should be equipped with both upward-looking and sidescan sonar, to map both the ice thickness and the bottom topography, the latter in order to detect evidence of bottom melt effects (Wadhams 1997) or pressure ridge disintegration (Schramm et al. 2000) that may be contributing to the observed ice loss.

As a step in the development of such techniques, we have succeeded in obtaining sidescan sonar images from the underside of Arctic sea ice using an AUV, the first time that this has been done in the Arctic and also the first measurements done during a polar winter. The measurements were carried out in February 2002 from R/V Lance (Norsk Polarinstitutt) using a Maridan Martin 150 vehicle in the East Greenland pack ice at 73°00′N, 11°47′W, as part of a research cruise to study Greenland Sea convection under the European Union (EU) CONVECTION program. In this paper we show some examples of the results obtained.

2. The vehicle and the experiments

The experiments were carried out within a mixture of first-year and multiyear ice in the winter marginal ice zone (MIZ) of the East Greenland Current (Fig. 1), where the ice was broken into floes by wave action but where fresh ice was growing in the interstices between floes in the form of frazil and pancake ice (the types of new ice that form in turbulent conditions).

The AUV used for the field campaign was the Maridan Martin 150. The Martin had a successful track record in inshore surveys but until this time had not been used in the open ocean. Its specifications were as follows: length 4.5 m; beam, including hydroplanes, 2.0 m; height 0.6 m; dry weight 900 kg; and operational depth 150 m.

One of the key criteria of an AUV is its navigational accuracy. The navigation and positioning system used on the Martin 150 was the MARPOS system. The core is an inertial navigation system using a ring laser gyro, coupled to a Doppler velocity log with a Trimble differential global positioning for surface fixes. In addition, the system is fed with information from a DigiQuartz pressure sensor to maintain accurate depth. The expected horizontal error from this system is 0.1% of the track distance. However, accurate navigation of an AUV where bottom tracking is not possible, or where the surface being followed is moving (i.e., sea ice), is a major challenge. This was the case for these runs; however, the accuracy achieved was better than 25 m at all times during the submerged missions and typically 15 m when real and programmed trajectories are compared (Maridan 2003, personal communication).

The scientific payload included a Tritech International SeaKing 675-kHz sidescan sonar (with two transducers, located on the port and starboard sides of the AUV), a conductivity–temperature–depth sensor (CTD) and an acoustic Doppler current profiler (ADCP). The SeaKing 675 has a ping rate of 5–7 Hz, a pulse length of 50– 200 μs, a 3-dB swath width of 50°, and an along-track beamwidth of 0.7°.

The AUV was housed in an insulated 20-ft container located on the foredeck of the ship. The container had access to heat and power and acted as the AUV storage and maintenance facility during survey. The AUV was deployed and recovered by a crane. The AUV can be seen in Figs. 2a and 2b. During launching and buoyancy trials the weather was uniformly bad with very high winds, and superficial damage to the vehicle was caused during launch and recovery, necessitating repairs. On the day of the successful data-gathering mission (27 February) the air temperature was −12.5°C, and wind speed had moderated to 11 m s−1 from the north.

The depth of the first run was 20 m at a speed of 1.2 m s−1. This depth was chosen to avoid the deepest ice in the region as estimated using the statistics of ice-thickness profiles from earlier submarine missions (Wadhams 1992). After completing the mission data were sent back via a high-frequency radio link to the ship. Via this link the AUV was reprogrammed for the second run, eliminating the need for it to be recovered between runs. A scan through the data revealed that ice in the region was less than 5 m thick, and thus the second run was performed at a shallower depth of 10 m and the same speed. Figure 1 shows the geometry of the two runs and the location of the operation: in run 1 the two parallel legs were 30 m apart while in run 2 they were 80 m apart. The total track length was 4.6 km.

3. Deriving ice draft from sidescan sonar measurements

Under-ice sidescan sonar was first used under ice in the Arctic on a submarine in 1976 (Wadhams 1978) and subsequently in 1987 (Wadhams 1988; Wadhams and Martin 1990), in both cases using equipment fitted by the experimenter. However, sidescan sonar is a well-proven acoustic technique for mapping features on the seafloor (e.g. Blondel and Murton 1997; Medwin and Clay 1998). It works (Seatronics 2002) by transmitting a fan-shaped beam of acoustic energy outward from port and starboard transducers so as to sweep the seabed in a line-scan fashion as the vehicle carrying the transducers (submarine, ship, or towfish) advances. As the energy radiates outward a proportion is reflected from objects intercepted at different distances from the sonar. The intensity of the returning energy or echo is a function of the shape and density of the objects encountered. Furthermore, a target with relief will result in an absence of sound, or shadows, behind the object, just as shining a beam of light at a low angle along the ground will create shadows behind objects that it hits. This intensity variation manifests itself as tonal variations, with light and dark portions representing strong and weak echoes, respectively. After each pulse the sonar waits for a predetermined time for the sound to be reflected back before transmitting another pulse. It is important to note that sidescan, unlike multibeam, has only one receiver per ship side. Thus, declination must be inferred from target distance only, using the assumption that the target field is essentially flat and horizontal.

Although the raw swath from the sidescan appears to be a real representation of the target area, it is in fact distorted, with the greatest distortion occurring in the near range of the record. This is due to the geometry of the system, as the initial return to the sidescan transducer is approximately vertical and the return from far range is approaching horizontal.

Producing a true two-dimensional image from sidescan can be complex; however, by assuming that the first sonar return is from ice immediately above the AUV, the process can be simplified. This assumption may not be valid in all cases, as it assumes that the ice is generally flat. It is conceivable that thicker ice, that is, a keel, may exist offtrack, yet with a travel time for the sonar pulse that is shorter and thus returns earlier than from ice that is closer to the vehicle's track. With this proviso, a sequence of processing techniques was applied to the raw sidescan data in order to obtain a corrected two-dimensional image. They were the removal of the water column, slant-range correction, and velocity correction. As there were only very small excursions in AUV pitch and roll throughout the missions, no corrections were made in this instance.

a. Removal of the water column

The water column offset is the distance between the AUV and the bottom of the ice. In order to produce an image swath of the underside of the ice this must be removed before the port and starboard sidescan sonar swaths are married together. Sea ice is acoustically highly reflective and leaves a bright, well-defined image on a sidescan record; thus, the interface between the water and the bottom of the ice is generally very pronounced (Fig. 3). However, noise within the return signal can make the automatic detection of the interface difficult and in some cases can lead to an erroneous detection of the ice–water interface. Noise in the sonar record manifested itself as either a bright along-track streak at the start of the record, due to return signal from the sonar–water interface, or as random cross-track strips throughout the record, possibly due to noise in the water column (e.g., ship's noise). The return signal from the sonar–water interface only affected the first few bins and thus was easily removed. Random noise (the horizontal streaks) was removed by a Butterworth filter. An intensity threshold was chosen to distinguish the water– ice interface and thus identify into which bin it fell. Figure 3b shows the ice–water interface line, displayed in red, superimposed over the raw data.

b. Deriving ice draft

In our case the intensity of the sonar returns was delivered in bins, with each bin representing an increase in the slant range of 0.1 m. In order to obtain an ice draft profile from the raw sidescan sonar data the location of the water–ice interface must first be correctly located. Once located, the bin that corresponds to the interface can be identified, and the distance from the AUV to the ice bottom surface obtained. The depth of the AUV is known at all times from the time-stamped CTD log. However, the AUV's internal microprocessor recorded the CTD and positional data at a slightly lower frequency than the sidescan data. In order not to lose the high definition of the sidescan images, the CTD and navigational data were interpolated to the same frequency as the sidescan, allowing each time stamp to have navigational, oceanographic, and sidescan data associated with it. Finally, the ice draft was calculated by subtracting the vertical distance from AUV to bottom of the ice from the AUV depth at the time. A constant was further subtracted from these measurements to allow for the difference between the vehicle location of the CTD and the sidescan. This method was applied to each sonar, and thus ice draft measurements for both the port and starboard sides of the AUV were obtained. Owing to the lateral separation of the two sonar units of 1.5 m, each sonar does not see the same point on the underside of the ice; thus, the ice draft from each channel is slightly different, and a 1.5-m zone of missing data exists in the middle.

c. Slant-range correction

The slant range is defined as the straight-line distance from the sidescan transducer to the point of reflection on the target, assumed to be the hypotenuse of a triangle defined by the AUV, the water–ice interface directly above it and the reflection point. Through simple trigonometry the “real” horizontal distance between bins can be calculated, making the assumption that the draft variations are small compared to the vehicle depth. Deep keels appear closer to the centerline than their real position, so the full image is subject to local distortions.

d. Velocity correction

The last stage in the production of a correct 2D image with a 1:1 aspect ratio was to correct for inconsistencies in the along-track velocity of the AUV. Navigational data from the AUV were used to calculate the distance traveled between successive pings of the sonar using the World Geodetic System 1984 (WGS 84) ellipsoid. With each ping of the sonar now associated with a distance from the start of the run, a true visualization of the underside of the ice could be obtained.

4. Results from sidescan

We now show some examples of the imagery obtained, along with corresponding draft profiles of the portion of track that lies along the centerline of the image.

Figure 4 is a 100-m-long sidescan section starting at 700 m into the first run of the AUV. It reveals a close-packed array of floes, the majority with smooth featureless bottoms. Since the almond-shaped floe (40 m across) that crosses the centerline of the survey track is 2.3 m thick, it is therefore likely to be multiyear ice, as first-year ice at this latitude in the Greenland Sea is usually around 1–1.5 m thick. The smooth homogeneous bottom of this and other floes in the images shows that considerable bottom melt has been going on, since in the Arctic Basin itself multiyear ice has a very rugged underside morphology of bulges and depressions, which enables it to be readily distinguished from smooth-bottomed first-year ice on sidescan imagery (Wadhams 1988). Many of the other floes in the image have considerable thickness, as shown by the width of the shadow zone on the side distant from the sidescan transducer, yet are only a few meters in diameter, sometimes 2–3 m. Such thick, but small, ice cakes are called brash ice and are the product of the extreme storm conditions to which this part of the ice field (and the R/V Lance) had been subjected for the preceding 3 weeks, which caused floes to collide vigorously and break into a variety of shapes and sizes. Apart from the deep floes and brash fragments, there is a light gray background with no apparent morphological contrast that surface observations showed to be frazil ice, that is, a suspension of new ice crystals in water, growing rapidly in the cold air but unable to form a continuous sheet on account of the high turbulence. In the lee of the floes there are also occasional black patches of open water. The black–white difference here, of course, is actually a difference between the high sonar backscatter of frazil ice (light) and the low backscatter of open water (dark), as frazil ice has a higher acoustic backscatter.

Figure 5 is again from the first run of the AUV and shows a more open part of the ice field, with much open water in the upper section of the image. Within the open water are narrow Langmuir streaks of newly forming frazil ice. A thick ice floe lying off the centerline near the top is distinguished by its strong (white) nearside reflections. Of interest is the floe in the center at the bottom, which has structure on its underside. Because it is offtrack we can only estimate its thickness from the length of its farside shadow, which gives an estimate of 2 m, suggesting that it is a multiyear floe. One possibility is that this is a floe that has not been subjected to much melt and retains its characteristic ruggedness. A more likely possibility is that the protuberances on the bottom are actually brash fragments that have been rafted under the floe in the violent conditions that prevailed on preceding days; visual observation revealed many floes with brash fragments on the surface that had been rafted over the floe by waves breaking against them.

Figure 6 comes from the second run of the AUV and shows a 110-m section of the sidescan imagery starting from 1330 m. This image shows heavy ice conditions with little or no open water present. In general the AUV ran under more compact ice conditions during the second run. The three angular multiyear floes, each up to 2.5 m thick, lying along the track are likely to have originated from the same floe that broke up because of collisions or wave action. Their undersides have an undulating structure but are basically smooth. Surrounding these floes is a network of smaller floes and brash fragments embedded in frazil ice.

A 200-m section of sidescan imagery from the second run can be seen in Fig. 7. This again shows very high concentrations of ice, with all ice types, from frazil to multiyear floes, being present. Toward the left-hand side is a very thick floe (around 5 m) casting a long shadow. This floe can easily be seen in the draft profile and is possibly rafted under another floe or a thick multiyear floe. On the far right are two other very thick floes (lying just off centerline to left and right) with some structure visible on the underside of the lower floe, suggesting a multiyear floe where the bottom topography has not yet been completely made smooth by the heat flux associated with the high ice–water velocities in this turbulent environment.

Figure 8 shows a 100-m section from run 2. Of interest is the floe in the center of the image, which has structure on its underside. This floe is almost 3 m thick where it crosses the centerline, so it is clearly a multiyear floe; however, when including the structure adhered to the bottom of the floe it has a draft of almost 6 m (estimated from the shadow that it casts). The shape and size of the protrusion on the bottom suggests that a brash fragment has been rafted under the larger floe. In the upper-right sector of the image we see Langmuir streaks of frazil ice surrounded by open water. On closer examination, however, there is a slight pattern to the water that indicates low concentrations of frazil at the surface of the water; that is, frazil ice is just beginning to form in this region.

5. Results for ice draft

Ice draft profiles, of which Figs. 4–8 show examples, were used to generate probability density functions (PDFs) of draft. Figure 9 shows the probability density of ice draft in 0.5-m bins for each of the two runs, with data from port and starboard sensors combined. The data are taken over the whole of each run (1700 m for run 1, 2900 m for run 2). There are some differences between the PDFs, but on the whole there is good agreement between both missions. The large peak at 0–0.5 m corresponds predominantly to the open water and frazil ice that lay between the floes, while the rest of the PDFs show a general tailing off toward the larger ice drafts. The modal draft for the melting polar ice floes encountered is 1–1.5 m, and the maximum draft is 5 m, corresponding to a floe in Fig. 7. The mean drafts (including open water) are 0.75 and 0.87 m for each run, respectively. The closest comparable data from the winter months comprised an upward sonar profile collected in May 1987 by HMS Superb (Wadhams 1992) from the East Greenland Current; the results for 72°– 74°N (220 km of data), also shown in Fig. 9, agree remarkably well. The exception is that the deepest ice observed in 1987 was 7 m (72°–73°N) and 8 m (73°– 74°N), but this may simply be a statistical artifact of the short record length in 2002.

Probability density functions of shorter stretches of data, in 0.1-m bins, show the impact of individual ice floes. Figure 10a shows the PDF from the combined port and starboard sensors at 700–800 m along run 2, where the preferred drafts are clearly separated into 0.7– 0.8, 1.5–1.6, and 2.8–3.0 m, respectively. These drafts correspond to individual floes of different ages and highlight the mixture of ice types found within the MIZ. Figure 10b shows the PDF, again from the combination of the port and starboard sensors, at 1700–1800 m along run 2. It is evident that the majority of ice is first-year floes, due to its shallow draft; however, the influence of the 5-m ice block (seen in Fig. 7) is clearly seen.

6. Conclusions

We have demonstrated that it is possible to perform high quality AUV surveys in the extreme polar environment in winter and obtain data that are comparable to those taken by manned submarines. It is particularly reassuring that we were able to visualize all forms of sea ice from frazil, never seen in a sidescan image before, to multiyear ice.

Future AUVs for ice-bottom imaging should clearly be equipped with multibeam sonar for complete quantitative swath mapping of ice draft, or at the very least a combination of simple sidescan and narrowbeam upward-looking sonar.

It is hoped that this successful ice-profiling mission will be a precursor to larger-scale missions with longer-range AUVs that will develop into a major monitoring effort for Arctic sea ice changes. If ice velocity is known (e.g., from drifters or successive satellite images), ice draft distribution itself, as measured by the AUV, gives mass fluxes. Also, ice thickness results from future AUV runs can be used to validate freeboard estimates from the new altimeters (laser and radar) aboard IceSat and Cryosat, in order to allow mean ice thickness to be estimated throughout the Arctic.

The AUV has many advantages, notably the high resolution that is possible by sailing close to the ice bottom, a benefit that manned submarines cannot enjoy for safety reasons, and the possibility of a closely controlled tight or overlapping grid of imaging tracks. The main drawback of the present AUV is lack of range.

Acknowledgments

We are grateful to Bo Krogh, the Maridan senior surveyor on board, for the deployment of the vehicle. The work was supported by the European Union under Contract EVK2-2000-00058 for the CONVECTION program.

REFERENCES

  • Bjørgo, E., Johannessen O. M. , and Miles M. W. , 1997: Analysis of merged SMMR–SSMI time series of Arctic and Antarctic sea ice parameters 1978–1995. Geophys. Res. Lett, 24 , 413416.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Blondel, P., and Murton B. J. , 1997: Handbook of Seafloor Sonar Imagery. Wiley, 314 pp.

  • Medwin, H., and Clay C. S. , 1998: Fundamentals of Acoustical Oceanography. Academic Press, 712 pp.

  • Rothrock, D. A., Yu Y. , and Maykut G. A. , 1999: Thinning of the Arctic sea-ice cover. Geophys. Res. Lett, 26 , 34693472.

  • Schramm, J. L., Flato G. M. , and Curry J. C. , 2000: Toward the modeling of enhanced basal melting in ridge keels. J. Geophys. Res, 105 (C6) 1408114092.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Seatronics, cited 2002: Tritech ROV Sidescan Sonar. Seatronics datasheet. [Available online at http://seatronics-group.com/geophys/ztriside.htm.].

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., and Wallace J. M. , 1998: The Arctic Oscillation signature in the wintertime geopotential height and temperature fields. Geophys. Res. Lett, 25 , 12971300.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wadhams, P., 1978: Sidescan sonar imagery of sea ice in the Arctic Ocean. Can. J. Remote Sens, 4 , 161173.

  • Wadhams, P., 1988: The underside of Arctic sea ice imaged by sidescan sonar. Nature, 333 (6169) 161164.

  • Wadhams, P., 1992: Sea ice thickness distribution in the Greenland Sea and Eurasian Basin, May 1987. J. Geophys. Res, 102 (C4) 53315348.

    • Search Google Scholar
    • Export Citation
  • Wadhams, P., 1997: Variability of Arctic sea ice thickness—Statistical significance and its relationship to heat flux. Operational Oceanography. The Challenge for European Co-operation,. J. H. Stel et al., Eds., Oceanography Series, Vol. 62, Elsevier, 368–384.

    • Search Google Scholar
    • Export Citation
  • Wadhams, P., 2000: Ice in the Ocean. Taylor and Francis, 351 pp.

  • Wadhams, P., and Davis N. R. , 2000: Further evidence for ice thinning in the Arctic Ocean. Geophys. Res. Lett, 27 , 39733975.

  • Wadhams, P., and Martin S. , 1990: Processes determining the bottom topography of multiyear Arctic sea ice. Sea Ice Properties and Processes Monogr. 90-1, U.S. Army Cold Regions Research and Engineering Laboratory, 136–141.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Location map showing the Greenland Sea region, with Greenland to the west and Spitsbergen to the northeast, and the ice edge from passive microwave data (north–south line). The geometry of the two ice-profiling runs can be seen in the insert

Citation: Journal of Atmospheric and Oceanic Technology 21, 9; 10.1175/1520-0426(2004)021<1462:SSIOTW>2.0.CO;2

Fig. 2.
Fig. 2.

(a) An AUV surfacing in the marginal ice zone of the East Greenland Current. All ice types were present from newly formed ice to multiyear floes. (b) Recovery of the AUV by starboard-side crane. A zodiac was lowered into the water to attach a safety line to the AUV

Citation: Journal of Atmospheric and Oceanic Technology 21, 9; 10.1175/1520-0426(2004)021<1462:SSIOTW>2.0.CO;2

Fig. 3.
Fig. 3.

(a) Subset of raw data from port sidescan sonar. The dark region, to the left of the picture, represents the water column above the AUV. (b) Image shown in Fig. 1a but with the water–ice interface (red line) detected using the method described within the text

Citation: Journal of Atmospheric and Oceanic Technology 21, 9; 10.1175/1520-0426(2004)021<1462:SSIOTW>2.0.CO;2

Fig. 4.
Fig. 4.

(top) A 100-m section of sidescan from mission 1 showing (a) a large multiyear floe surrounded by a network of smaller floes and brash fragments embedded in frazil ice. (bottom) The ice draft as seen directly above the port and starboard sidescan sensors

Citation: Journal of Atmospheric and Oceanic Technology 21, 9; 10.1175/1520-0426(2004)021<1462:SSIOTW>2.0.CO;2

Fig. 5.
Fig. 5.

(top) Section of sidescan from mission 1 showing a more open part of the ice field, with much open water, and (a) narrow Langmuir frazil streaks in the upper section of the plot and (b) a floe in the center at the bottom that has structure on its underside. (bottom) The ice draft as seen directly above the port and starboard sidescan sensors

Citation: Journal of Atmospheric and Oceanic Technology 21, 9; 10.1175/1520-0426(2004)021<1462:SSIOTW>2.0.CO;2

Fig. 6.
Fig. 6.

(top) Section of sidescan from mission 2 showing a high concentration of all ice types: frazil, brash/pancake, and floes. (a), (b), (c) Three angular multiyear floes, each up to 2.5 m thick can be seen in the central region of the image. No open water is visible in the image. (bottom) The ice draft as seen directly above the port and starboard sidescan sensors

Citation: Journal of Atmospheric and Oceanic Technology 21, 9; 10.1175/1520-0426(2004)021<1462:SSIOTW>2.0.CO;2

Fig. 7.
Fig. 7.

(top) A 200-m section of sidescan from mission 2 showing (a) a very thick floe (about 5 m) toward the left-hand side of image. Again the image is of high concentrations of ice (all types) with little or no open water present. (bottom) The ice draft as seen directly above the port and starboard sidescan sensors

Citation: Journal of Atmospheric and Oceanic Technology 21, 9; 10.1175/1520-0426(2004)021<1462:SSIOTW>2.0.CO;2

Fig. 8.
Fig. 8.

(top) Section of sidescan image from run 2 showing both open water (top right) and closed pack conditions (bottom). (a) A floe almost 3 m thick includes a structure adhered to the bottom. The shape and size of the protrusion on the bottom suggests that a brash fragment has been rafted under the larger floe. Shown in the upper-right sector of the image are (b) Langmuir streaks of frazil ice surrounded by open water. (bottom) The ice draft as seen directly above the port and starboard sidescan sensors

Citation: Journal of Atmospheric and Oceanic Technology 21, 9; 10.1175/1520-0426(2004)021<1462:SSIOTW>2.0.CO;2

Fig. 9.
Fig. 9.

Probability density functions of ice draft along the centerlines of (a) runs 1 and (b) 2 for combined port and starboard channels are shown in black. Also shown, in gray, are draft distributions obtained in May 1987 from HMS Superb from the latitude range 72°– 74°N

Citation: Journal of Atmospheric and Oceanic Technology 21, 9; 10.1175/1520-0426(2004)021<1462:SSIOTW>2.0.CO;2

Fig. 10.
Fig. 10.

(a) Probability density functions of ice draft (in 0.1-m bins) along the centerline for combined port and starboard channels between 700 and 800 m along run 2. Individual floes of different ages correspond to different peaks. (b) Probability density functions of ice draft (in 0.1-m bins) along the centerline for combined port and starboard channels between 1700 and 1800 m along run 2. The influence of the rafted ice block surrounded by thinner first-year floes (0.5–1.8 m) is clearly visible

Citation: Journal of Atmospheric and Oceanic Technology 21, 9; 10.1175/1520-0426(2004)021<1462:SSIOTW>2.0.CO;2

Save
  • Bjørgo, E., Johannessen O. M. , and Miles M. W. , 1997: Analysis of merged SMMR–SSMI time series of Arctic and Antarctic sea ice parameters 1978–1995. Geophys. Res. Lett, 24 , 413416.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Blondel, P., and Murton B. J. , 1997: Handbook of Seafloor Sonar Imagery. Wiley, 314 pp.

  • Medwin, H., and Clay C. S. , 1998: Fundamentals of Acoustical Oceanography. Academic Press, 712 pp.

  • Rothrock, D. A., Yu Y. , and Maykut G. A. , 1999: Thinning of the Arctic sea-ice cover. Geophys. Res. Lett, 26 , 34693472.

  • Schramm, J. L., Flato G. M. , and Curry J. C. , 2000: Toward the modeling of enhanced basal melting in ridge keels. J. Geophys. Res, 105 (C6) 1408114092.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Seatronics, cited 2002: Tritech ROV Sidescan Sonar. Seatronics datasheet. [Available online at http://seatronics-group.com/geophys/ztriside.htm.].

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., and Wallace J. M. , 1998: The Arctic Oscillation signature in the wintertime geopotential height and temperature fields. Geophys. Res. Lett, 25 , 12971300.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wadhams, P., 1978: Sidescan sonar imagery of sea ice in the Arctic Ocean. Can. J. Remote Sens, 4 , 161173.

  • Wadhams, P., 1988: The underside of Arctic sea ice imaged by sidescan sonar. Nature, 333 (6169) 161164.

  • Wadhams, P., 1992: Sea ice thickness distribution in the Greenland Sea and Eurasian Basin, May 1987. J. Geophys. Res, 102 (C4) 53315348.

    • Search Google Scholar
    • Export Citation
  • Wadhams, P., 1997: Variability of Arctic sea ice thickness—Statistical significance and its relationship to heat flux. Operational Oceanography. The Challenge for European Co-operation,. J. H. Stel et al., Eds., Oceanography Series, Vol. 62, Elsevier, 368–384.

    • Search Google Scholar
    • Export Citation
  • Wadhams, P., 2000: Ice in the Ocean. Taylor and Francis, 351 pp.

  • Wadhams, P., and Davis N. R. , 2000: Further evidence for ice thinning in the Arctic Ocean. Geophys. Res. Lett, 27 , 39733975.

  • Wadhams, P., and Martin S. , 1990: Processes determining the bottom topography of multiyear Arctic sea ice. Sea Ice Properties and Processes Monogr. 90-1, U.S. Army Cold Regions Research and Engineering Laboratory, 136–141.

    • Search Google Scholar
    • Export Citation
  • Fig. 1.

    Location map showing the Greenland Sea region, with Greenland to the west and Spitsbergen to the northeast, and the ice edge from passive microwave data (north–south line). The geometry of the two ice-profiling runs can be seen in the insert

  • Fig. 2.

    (a) An AUV surfacing in the marginal ice zone of the East Greenland Current. All ice types were present from newly formed ice to multiyear floes. (b) Recovery of the AUV by starboard-side crane. A zodiac was lowered into the water to attach a safety line to the AUV

  • Fig. 3.

    (a) Subset of raw data from port sidescan sonar. The dark region, to the left of the picture, represents the water column above the AUV. (b) Image shown in Fig. 1a but with the water–ice interface (red line) detected using the method described within the text

  • Fig. 4.

    (top) A 100-m section of sidescan from mission 1 showing (a) a large multiyear floe surrounded by a network of smaller floes and brash fragments embedded in frazil ice. (bottom) The ice draft as seen directly above the port and starboard sidescan sensors

  • Fig. 5.

    (top) Section of sidescan from mission 1 showing a more open part of the ice field, with much open water, and (a) narrow Langmuir frazil streaks in the upper section of the plot and (b) a floe in the center at the bottom that has structure on its underside. (bottom) The ice draft as seen directly above the port and starboard sidescan sensors

  • Fig. 6.

    (top) Section of sidescan from mission 2 showing a high concentration of all ice types: frazil, brash/pancake, and floes. (a), (b), (c) Three angular multiyear floes, each up to 2.5 m thick can be seen in the central region of the image. No open water is visible in the image. (bottom) The ice draft as seen directly above the port and starboard sidescan sensors

  • Fig. 7.

    (top) A 200-m section of sidescan from mission 2 showing (a) a very thick floe (about 5 m) toward the left-hand side of image. Again the image is of high concentrations of ice (all types) with little or no open water present. (bottom) The ice draft as seen directly above the port and starboard sidescan sensors

  • Fig. 8.

    (top) Section of sidescan image from run 2 showing both open water (top right) and closed pack conditions (bottom). (a) A floe almost 3 m thick includes a structure adhered to the bottom. The shape and size of the protrusion on the bottom suggests that a brash fragment has been rafted under the larger floe. Shown in the upper-right sector of the image are (b) Langmuir streaks of frazil ice surrounded by open water. (bottom) The ice draft as seen directly above the port and starboard sidescan sensors

  • Fig. 9.

    Probability density functions of ice draft along the centerlines of (a) runs 1 and (b) 2 for combined port and starboard channels are shown in black. Also shown, in gray, are draft distributions obtained in May 1987 from HMS Superb from the latitude range 72°– 74°N

  • Fig. 10.

    (a) Probability density functions of ice draft (in 0.1-m bins) along the centerline for combined port and starboard channels between 700 and 800 m along run 2. Individual floes of different ages correspond to different peaks. (b) Probability density functions of ice draft (in 0.1-m bins) along the centerline for combined port and starboard channels between 1700 and 1800 m along run 2. The influence of the rafted ice block surrounded by thinner first-year floes (0.5–1.8 m) is clearly visible

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 634 176 10
PDF Downloads 309 88 1